Be the origin of the scale what it may, its prac tical construction is well known. The space be tween freezing and boiling water is divided into 180°, and these degrees being carried down, the zero is placed 32° below freezing water, at a de gree of cold, observed by Fahrenheit (though some insist by Ronaur) in Iceland, in 1709. The scale of the Mercurial Termometer has since been vari ously divided, but except two, these graduations have fallen into disuse. The first of these is usual ly called Reaumur's, though we shall presently show that it differs practically from his, and was really applied by Deluc: it divides the space be tween freezing and boiling water into 80', whence the length of a degree is w or of Fahrenheit. Its zero is at the freezing point. The second is the scale of Celsius, the Swedish philosopher, though revived during the frenzy of the French revolution ists for decimal measures, under the title of Centi grade: it has 100° between freezing and boiling water, and its scale begins at the former point; the principal defects of this and the preceding gradua tion are, that the degrees are too large, requiring the application of fractions, and that negative quan tities are perpetually occurring, both of which are in a great measure avoided in the less natural scale of Fahrenheit. The relative values of the degrees on these scales are expressed by the following for mula, in which the degree on each scale is indicat ed by its initial letter.
A very simple and useful rule may be added for the conversion of the Centigrade into Fahrenheit's scale. Double the number of degrees and subtract one tenth, and the result will be the number of de grees above the freezing point, to which add 32°. Thus, to convert 74° Centigrade into Fahrenheit, we say, 74 X 2 = 148 and 148 -- 14.8 + 32° = 165.2.
In De Fisle's scale, the whole mercury at boiling heat is divided into 10,000 parts, each of which forms a degree counted downwards, so that the freezing point stands at 150°; a very unnatural graduation, but which is still used in Russia. The only other scale we shall notice, is one distinguish ed as being perhaps the most philosophical and convenient yet proposed, though it has never come into use; it is that of the late Dr. Murray of Edin burgh; he divides the whole range of mercury from freezing to boiling into 100 parts, whence if we as sume — 40 and + 655 Fahr. for those points, freezing water will be at and boiling water at 347°. Here the degrees would be of a very con venient length, (about ith of those of Fahrenheit) and negative quantities could hardly ever be re quired.' Such is an account of the mercurial thermome ter to the present day, for it has hitherto under gone no alterations of consequence in its simple form. We must now, in order, say a few words upon the real thermometer of Reaumur, which was constructed with spirit of wine, though in point of date it followed Fahrenheit's.
Reaumur conceived that lie had ascertained that 1000 parts of strong spirit of wine expanded to 1087 at the heat of boiling water, lie therefore diluted the fluid till he found the dilatation reduced to exactly 80 parts, which formed his degrees. Now this being a tentative process, ought by all means to be excluded from the practical construction of thermometers, and besides, all attempts which were then made to graduate upon the principles of abso lute fractional expansion were fruitless in a philo sophical point of view, for the expansion of the glass in which the fluid was contained, was always overlooked, as well as the inequalities of the dilata tion of the substance itself. But the great error of
Reaumur was in the determination of his fixed points, for since he never allowed the spirit to boil in the tube at the upper extremity of the scale, in stead of being 212 Fahrenheit, it was only about 175, as Martine has justly shown; the bulbs of his thermometers were likewise so large, that the freez ing point was about 2° Fahrenheit too high. This thermometer was therefore a most imperfect in strument, and now wholly abandoned; we think it therefore unnecessary to enter upon it more fully, but refer for a full description of it to the work of Delue,f in whose time the true value of Reaumur's degrees was of some importance, as the instrument had been much employed in France. That author afterwards applied the octogesimal division to the mercurial thermometer, which, as we have already observed, usually goes under the name of Reaumur's scale, for the conversion of which, into Fahren heit's, we have already given directions.
Two more modern thermometers remain to be noticed. The expansion of metallic substances from its small amount, and the degree of heat to which it might be applied, early pointed them out as suitable indicators of intense temperatures, and under the name of pyrometers they were employ ed, with various modifications, during the last cen tury (See PYROMETER.) The beautiful and refined application of the unequal expansibility of metals to the formation of a compensation pendulum, and especially the still more delicate contrivance for the compensation of the balance of a chronometer, probably suggested to Messrs. Breguet, the dis tinguished Parisian artists, the possibility of ren dering the minute variation of metallic expansion sufficiently visible for ordinary thermometric pur poses. Their instruments are of two descriptions. One in the form of a watch is shown at Plate DXXIV, Figs. 5 and 6, where, in the former, which represents the interior, the expanding metallic bar a a a a forms a double curve; it is composed of two lamina of steel and brass firmly united, the excess of the dilatability of the latter above the former causing the curve to expand and contract, thus acting upon the lever d, the opposite and longer arm of which consists of an arch-head, with fine teeth upon its edge, which act upon the small wheel upon the axis of which the needle which traverses the dial, shown in Fig. 6, is fixed. A fine gold wire spring acts upon the wheel, so as to keep the lever d always in contact with the extrem ity of the expanding bar. The most modern and beautiful form of the instrument is shown at Fig. 7, in which the metals employed are platina and silver, with a slip of gold interposed, as being of intermediate expansibility, the whole is flattened to an extreme tenuity and then formed into a spiral spring, to the lowest turn of which is attached an index which traverses a horizontal dial. This in strument is far superior to the last, and is alto gether one of the most elegant philosophical in struments we ever saw. We had lately an oppor tunity of inspecting a number of them at M. Bre guet's own manufactory.